MUSICによる近距離音源の高分解能位置推定法に関する研究

－センサ間の振幅・位相偏差とその補償処理が位置推定性能に及ぼす影響－

抄録

A MUSIC (MUltiple Slgnal Classification) is super resolution and high accuracy measurement algorithm for location finding of the signal sources of the received signals using the array sensors. However, it is known that the location finding error remarkably increases when the array receiver has amplitude and phase deviations between sensors coming and estimating the signal sources from far field environment. This paper describes the MUSIC characteristics for the near field sound sources when the receiver has sensor deviations. And also, this paper investigates compensated preprocessing algorithm for estimating and decreasing sensor deviations. For confirming the algorithm characteristics, the computer and experimental simulation are done, and it is clear the MUSIC and preprocessing algorithm are enable to properly estimate the location of sound source and sensors deviations.